Abstract:
This study addresses the critical challenge of optimizing energy management in a smart campus envi-
ronment through the integration of solar photovoltaic (PV) systems, energy storage, and dynamic load
balancing. Focusing on the Huye campus of the University of Rwanda, the project develops a robust
energy management system (EMS) that takes advantage of predictive analytics, stochastic and robust
optimization techniques, and real-time Model Predictive Control (MPC) to minimize grid reliance, re-
duce operational costs, and enhance sustainability. By analyzing historical energy consumption pat-
terns (2019–2023) and simulating scenarios such as sunny, cloudy, and grid outage conditions, the EMS
demonstrates a reduction of 60–92% in energy costs through prioritized solar utilization, demand re-
sponse (DR) strategies, and optimization of energy storage. A Decision Support Tool (DST) is integrated
to provide actionable insights, enabling campus managers to make data-driven decisions about energy
efficiency. Key results include an 81% reduction in grid dependency, a validated photovoltaic capacity
of 848 kWp, and a scalable framework applicable to educational institutions in solar-rich regions. Key
innovations include a scenario-based resilience framework for cloudy or rainy days and a Decision
Support Tool (DST) that uses data-driven insights, predictive analytics, and actionable recommenda-
tions to enhance system efficiency, reliability, and sustainability. Simulations demonstrate a 4.2-hour
backup during grid outages and a projected 6.2-year payback period for the 848-kWp PV system. The
work aligns with Rwanda’s National Energy Policy (2023) and offers a replicable model for regional
educational institutions.
